Tutorials > R Tidymodels Part 3: Classification

R Tidymodels Part 3: Classification

Date: 25 October 2025 | Posted by: 0nelove

R, Data Science, tidymodels, Machine Learning, Classification, Logistic Regression, KNN, Naive Bayes, Metrics, RStudio
What you'll learn
What classification is and how it differs from regression
How to interpret probabilities, odds, and log-odds in logistic regression
How logistic regression models are estimated using maximum likelihood
How to fit and interpret logistic regression models in tidymodels
What are the key classification metrics and how to calculate them
How to use accuracy, precision, recall, specificity, and F1 score
How to visualize and interpret ROC curves
How probability thresholds affect classification results
What happens when data are imbalanced and how to handle it
How to use balanced accuracy and advanced metrics for imbalanced data
How to tune probability thresholds and evaluate precision–recall curves
What resampling techniques exist for imbalanced data (up-sampling, down-sampling, SMOTE)
How to apply resampling techniques inside the tidymodels framework
How to train and tune KNN classifiers in tidymodels
What Bayes' theorem is and how it leads to the Naive Bayes classifier
How the Naive Bayes classifier works for discrete and continuous features
How to build and compare Naive Bayes models in tidymodels
Requirements
Finishing "R tidymodels part 2: Beyond linear regression" is strongly recommended.
Familiarity with the tidymodels framework and ML workflows covered in previous parts.
R and RStudio already installed on your computer.
Basic knowledge of statistics (concepts such as probability, regression and classification) is a plus.
Intermediate R knowledge and experience with tidyverse syntax are recommended.
If you are a complete beginner to programming or R, you may find this course challenging.
Interest in data science, machine learning, and classification modeling.
Curiosity about understanding probabilities, metrics, and model evaluation.
Interest in writing clean, efficient, and reproducible R code.
Please update R and its libraries if necessary. A list of versions (R and all R libraries used in the exercises) is provided at the end of each section.
Description
You've mastered regression modeling and explored advanced algorithms; now it's time to step into the world of classification.This course is designed for learners who want to build models that predict categories, not numbers, and who wish to understand the statistical and machine learning foundations behind them.What You'll LearnIn this course, you'll move from probability intuition to full-scale classification workflows using tidymodels, R's modern ecosystem for machine learning:Understand what classification is and how it differs from regressionLearn the logic of logistic regression and its link to probabilities, odds, and log-oddsFit and interpret logistic regression models using maximum likelihood estimationEvaluate models with accuracy, precision, recall, specificity, and F1 scoreVisualize model performance through ROC and AUC curvesAdjust probability thresholds and see their effect on predictionsHandle imbalanced data using balanced accuracy, threshold tuning, and advanced metricsApply resampling techniques such as upsampling, downsampling, and SMOTEBuild and tune K-nearest neighbors (KNN) classifiersExplore Naive Bayes as a probabilistic classifier for both numeric and text dataPreprocess text using textrecipes and create a simple spam-filtering modelCompare multiple classification models within the same tidymodels workflowWhy Take This Course?Classification problems are everywhere — from medical diagnostics and fraud detection to email filtering and customer segmentation.This course helps you understand how these models make decisions and how to evaluate them responsibly.You'll gain not only the technical skills to build classification models but also the intuition to select the right metric and interpret model behavior — all while keeping your work tidy, reproducible, and explainable.What You'll GetClear, structured explanations of classification theory and practiceStep-by-step modeling workflows in R and tidymodelsReal-world examples and visual explanations of metricsExercises and assignments with full solutionsAll code, datasets, and outputs providedLifetime access and updates
Who this course is for
Anyone interested in data science or machine learning
Learners who want to understand classification modeling in R
Anyone who wants to analyze and predict categorical outcomes
Data analysts and scientists who want to extend their tidymodels skills
Anyone curious about logistic regression, KNN, or Naive Bayes algorithms
Those who wish to learn how to evaluate models using real-world metrics
Professionals who work with imbalanced data or probability-based predictions
Students and researchers building classification models
R users aiming to deepen their modeling expertise within the tidyverse/tidymodels ecosystem
Data scientists who mainly use Python and want to extend their skills into R

See full article